Incosistent fuel table - fuel Pulsation?

Hello,

The Car: rx7 FD with stock porting, intake manifold and throttlebody with a Elite 2500 in VE Mode with TPS as Main Load source.

It has a FFE Rail Kit with 4 id1050x Injectors. The rails are connected in parallel to the pressure Regulator. The FPR ist installed remotly with relative long fuel Lines. About 80cm for each rail.

The deadtimes and flow tables are set in 3d with ID's calibration data. A fuel pressure sensor is installed. Injector pressure differential stays on point. Short pulsewidth adder table is also active

My Problem is that the VE table is incosistent in a lot of areas. Different opening angles of the throttle makes waves into the ve-table. Low throttle opening areas are incosistent with a lot of spikes an valleys. In the Cruise are is a big Spike (15%tps @ 1500rpm) 25 % Higher then the sourrounding cells. But with this map the lambda target is on point.

Can these errors be caused by fuel Pulsation? Or any other ideas?

Bastian

Hi Bastian,

Wankel engines require high gas speed to achieve decent combustion efficiency given the shape of the combustion chamber so you could expect some inconsistent or poor fuelling at low load/throttle openings but as you said fuel pulsation could exacerbate the issue. are you running dead head to the fuel rail? if not this combined with a pulsation damper can improve things especially when big injectors and small bore fuel rails are used.

Hi Scotty,

thanks for your reply.

I'm running a return fuel system. It has a separate fuel line for each of the rails. Each line has its own port on the regulator.

But between the rails and the FPR is about 80 cm of fuel line for each rail. So it's 1.6m of fuel lines. I think I have to fix this.

But can please have a look on the attached image. I don't understand where these two waves come from. These waves are almost over the whole table on the same RPM rows. 1500-2000 and 3000-3500. the TPS seems not to have a effect on it.

regards

Bastian

There is a notable drop in VE at low rpm when the secondary throttles open, it's a catch 22, you need more primary throttle for more air but then lose VE. To tune one properly you really need TPS and MAP load as you are using or a sensor input on secondary throttle and switching table to account for that of only speed density, especially if primary and secondary porting is different to get the best from the engine.

I intend to use the warm-up throttles to improve low rpm torque rather than just to restrict output as per factory when I tune mine.

Hi Slides,

thank for you reply.

I'm already using TPS as main Load Source with a MAP-Correction on it. This gave me also the best throttle response.

The idea to use the warm up plates to get better low end torque is great! I will give it a try!

How did you program the opening of the plates? Did you made two VE tables, one with throttle open and one with the throttle closed to see where the VE has the highest values and based on that you have programmed it?

regards

Bastian

Haven't hot there yet. I think mapping with them locked in either state to find the crossover would be easiest.

I installed a radium damper and changed the FPR from Aeromotive to a small FueLab FPR thats now mounted directly to the rail.

The spikes are a lot smaller now.

But whats still not working well is tuning the deceleration areas in the VEtable (TPS as Load Source) I still see VE values over 100. Maybe im doing the mapping of this area wrong.

How would you tune those low TPS areas vs high RPM. I was trying to do it with disabled decel cut and coast trough them

regards

Tune implies you have a target. What is the target? Maximum torque, maximum engine braking, best response if you go back to throttle at high RPMs?

Since the driver is not requesting any torque with the throttle closed -- I generally just move to the smallest possible fuel pulse if I'm not using Decel Fuel Cut Off.

My target was to reach my desired Lambda values from my lambda target table. But the VE numbers are unrealistically high for an area with closed throttle.

I'm just curious why I cant reach the Lambda target with VE numbers that are approximated with the cells around. E.g. 91% VE with throttle complety closed @ 4000 rpm to reach a lambda target of 0.95. I also want to understand why I have to add so much fuel while decelerating. I hear no misfiring that can cause lean reading.

I'm not sure that a VE of 91% with close throttle is adding much fuel at all, because the manifold pressure should be very low. Why don't you compare the injection pulse width at idle and this 4000/0% TPS to get a feel for how much fuel is really being injected.